11Imaging — Who to Scan and Why

FImaging is targeted, never routine

The default for tinnitus is not to scan. The overwhelming majority of patients have bilateral, non-pulsatile tinnitus with a symmetrical hearing loss, and imaging in that group has a vanishingly low yield while exposing patients to cost, incidental findings and, for CT, radiation. The professional guideline therefore recommends against routine imaging and reserves it for specific clinical features [2014].

The governing idea is simple: a red flag generates an anatomical question, and imaging is the tool that answers it. No red flag, no question, no scan.

TThe red flags that justify a scan

Four patterns move a patient from “no imaging” to “image deliberately.” Unilateral or asymmetric tinnitus, particularly with asymmetric sensorineural hearing loss, asks whether there is a retrocochlear lesion such as a vestibular schwannoma. Pulsatile tinnitus asks whether there is a vascular or structural source — a dural fistula, sinus stenosis or diverticulum, glomus tumour or aberrant vessel. Focal neurology — facial numbness or weakness, other cranial-nerve signs, ataxia — asks about a cerebellopontine-angle or brainstem lesion. An otoscopic mass, such as a retrotympanic red-blue lesion behind an intact drum, asks about a glomus tumour or vascular anomaly [2008].

Each flag maps to a different first-line modality, which is the subject of the next two modules: asymmetric SNHL/tinnitus to MRI of the IAC/brain, and pulsatile tinnitus to a vascular protocol tailored to whether the lesion is suspected to be venous, arterial or a temporal-bone mass.

CCost, yield and the harm of over-imaging

Targeting is justified by yield. In asymmetric hearing loss, the prevalence of an actionable retrocochlear lesion is low — on the order of a few percent — which is exactly why screening strategies were studied to triage who needs the scan [2004]. Scanning every tinnitus patient would mean hundreds of negative studies for each lesion found, plus a steady stream of incidental findings (white-matter spots, benign cysts) that generate anxiety and downstream tests.

The clinician’s job is to push patients into the right bin: reassure and observe the low-risk majority, and image decisively when a flag is present. Over-imaging is not a kindness — it is a source of harm and distraction from the counselling that most tinnitus patients actually need.

CBuilding the request and setting expectations

A useful imaging request names the clinical question, not just the body part: “asymmetric SNHL — exclude vestibular schwannoma” tells the radiologist which protocol to run far better than “MRI brain.” This is especially true in pulsatile tinnitus, where the suspected source (venous vs arterial vs mass) dictates whether venographic or arteriographic sequences are added [2024].

Counsel the patient honestly: most scans for tinnitus are normal, a normal scan is reassuring rather than a failure, and the purpose is to exclude a treatable structural cause — not to “find the tinnitus,” which imaging cannot do [2013].